Data Science & IA

Proactive Incident Management Model in a logistics company

The challenge

Logistics company was going through a period of constant growth, especially thanks to the e-commerce business. Both its customers and end-customers have been increasing their demands in terms of communication regarding the fulfillment of committed deadlines. It was costly for the company to identify incidents early, so it was not able to inform customers when there was a change in the delivery date. In order to solve this problem, it was decided to develop a minimum viable product (MVP) to anticipate problems in meeting delivery times by implementing a proactive incident management process.

The tool made it possible to anticipate problems in meeting delivery times by implementing a proactive incident management process.

The strategy

Based on the use of innovation methodologies (e.g. sprints) and a multidisciplinary work with different areas of the company, it was possible to generate in six weeks:

  • A data model based on the complete life of the shipment, which makes it possible to identify in a simple way in what state it is in.
  • Adjustments were made to the relevant processes, such as the requirement of the “puncture” at pick-up, which increases the traceability of the piece.
  • An alert system was created with automatic control of parts without movement, so as to be able to take timely action on each part.
  • A Customer Health dashboard was delivered, which identifies problems encountered with the addresses, e-mails and numbers provided by customers.
  • The withdrawal status was made available to the customer, thus reducing uncertainty and standard communication to report shipping incidents.
A data model based on the full life of the shipment

that makes it easy to identify in what state the product is in.

Adjustments to relevant processes

such as the requirement of "puncture" at removal, which allows to increase part traceability

Alert system with automatic control of non-moving parts

in order to be able to take timely action on each piece of equipment.

Customer Health Dashboard

in which identifies problems encountered with addresses, e-mails and numbers provided by customers

Availability of the withdrawal status to the customer

thus reducing your uncertainty and standardized communication for reporting shipping incidents.

The achievements

In the first days of measurement, promising results were delivered to the effect of the initially defined project objectives:

On average, 634 alerts are generated daily for parts with no movement, equivalent to 3.2% of parts

1,306 messages would be sent to customers indicating that the established date would not be met.

As for the information provided by customers, the gap is 59.8% invalid e-mails and 3.7% invalid numbers. We are still working to reduce this gap to 0%.

The withdrawal rate was raised by 5% for the Green Sale business.

156,285 messages were sent to customers to confirm receipt of the product, with a response rate of 7.6%.